Related Experiment Video
Updated: Aug 5, 2025
13:19
Enhanced Oil Recovery using a Combination of Biosurfactants
Published on: June 3, 2022
5.3K
Supervised deep learning-based paradigm to screen the enhanced oil recovery scenarios
Rakesh Kumar Pandey1, Asghar Gandomkar2, Behzad Vaferi3,4
1Department of Petroleum and Energy Studies, School of Engineering and Technology, DIT University, Dehradun, India.
Scientific Reports
|March 25, 2023
Summary
This study introduces a deep learning classifier to select the best enhanced oil recovery (EOR) method using reservoir properties. The novel model accurately identifies optimal EOR strategies, improving economic efficiency in oil production.
Area of Science:
- Petroleum Engineering
- Artificial Intelligence in Energy
- Reservoir Management
Background:
- Rising oil prices and dwindling reserves necessitate efficient Enhanced Oil Recovery (EOR) techniques.
- Optimizing EOR selection is crucial for economic viability and maximizing oil extraction.
- Existing methods for EOR selection can be complex and data-intensive.
Purpose of the Study:
- To develop a novel deep learning classifier for selecting the optimal EOR method.
- To base EOR selection on key reservoir rock and fluid properties, including depth, porosity, permeability, gravity, viscosity, and temperature.
- To create an efficient tool for screening EOR scenarios and reducing economic costs.
Main Methods:
- A hybrid deep learning model combining 1D convolutional neural network (CNN), long short-term memory (LSTM), and densely connected layers was designed.
- The genetic algorithm was employed for hyperparameter tuning of the deep learning classifier.
- The model was trained and validated on a dataset of 735 EOR projects across diverse reservoir types and countries.
Main Results:
- The deep learning classifier achieved high accuracy in identifying the best EOR methods, with overall classification accuracy reaching 96.82% for training, 84.31% for validation, and 82.61% for testing.
- The model demonstrated a low categorical cross-entropy of 0.1548, indicating effective classification of EOR techniques.
- Numerical and graphical analyses confirmed the reliability of the deep learning classifier in screening EOR scenarios.
Conclusions:
- The developed deep learning classifier is a robust and reliable tool for selecting the most suitable EOR method based on reservoir characteristics.
- This approach offers a powerful solution for optimizing EOR strategies, especially when field information is limited.
- The study highlights the potential of AI in enhancing efficiency and economic outcomes in the oil and gas industry.
